LETTER TO THE EDITOR
OCT angiography in idiopathic macular holes,
some methodological concerns
Received: 28 April 2017 /Accepted: 19 June 2017 /Published online: 1 July 2017
Springer-Verlag GmbH Germany 2017
Optical coherence tomography angiography (OCTA)
drives a lot of attention, as it enables exploration more thor-
oughly than ever on how the retinal and choroidal circulation
behave in health and disease. In their prospective study ,
Yufei Teng et al. compared the operated eye of 25 patients
with unilateral idiopathic macular hole (IMH) with the fellow
eye as well as with matched healthy controls. They measured
the choriocapillary flow area and parafoveal vessel density
and found them to be significantly smaller and lower in
IMH eyes than in non-affected eyes. They showed that those
values were significantly increased after vitrectomy. They also
examined five eyes with a lamellar hole and found no differ-
ence with the fellow healthy eye.
The authors infer from the findings that the parafoveolar
choroidal circulations are altered in IMH and can be restored
following closure of IMH. They conclude that the
choriocapillary circulation has a crucial role in macular holes.
The authors acknowledge some bias related to the imaging
technique. However, an important cause of artifact has not
been taken into account.
The presence of retinal cysts influences the capture of
vascular structures by en face OCT, including in the cho-
roidal planes [2–4]. It is true that the angiographic mode
relies on temporal signal changes within the tissue, but the
capture of deep vascular signals depends on the properties
of the overlying tissue .
Since intraretinal cysts are a common finding around IMH
[6–8], those changes and their effect on the way the vessels
appear at selected planes cannot be ignored when measuring
the choriocapillary flow area and parafoveal vessel density.
Indeed, Fig. 1 illustrates this phenomenon better than
any sentence . The central circle of 0.6 mm diameter,
corresponding to the area over which the neurosensorial
tissue is absent, appears as hyperreflective on the en face
OCT angiogram. The decrease in vascular density aound
this area correlates with the presence and dimensions of
the perifoveal cysts (well appreciated on the B-scan, c).
On A and B, the vascular signal is more attenuated on the
right than the left part of the en face images, as right cysts
are more substantial in the overlying retina.
The authors state that five eyes with a lamellar hole elim-
inate bias related to overlying tissues. This would be correct if
the bias were related to retinal thickness alone, but here the
structure of the overlying retina is relevant. As long as no
macular cysts complicate those cases, changes in the overly-
ing retina as a source of artifact have not been ruled out.
In this study, no significant correlation was found between
the duration of IMH and the choriocapillary flow area or
parafoveal vessel density. The duration of MH is a recognized
factor in IMH prognosis . Moreover, subfoveal choroidal
blood flow depends on the photoreceptor stimulation , se-
verely affected by IMH . A decrease in choriocapillary
flow area and parafoveal vessel density with duration of
IMH would, therefore, not be surprising. In this study, this
decrease may not have been detected because the so-called
choriocapillary signal actually reflects the presence or absence
of intraretinal cysts, that depend on anatomic closure, rather
than real changes in the choriocapillary flow.
The choriocapillary circulation may have a crucial role in
IMH. However, this assertion cannot be inferred from the
presented data. An elegant way to check the accuracy of the
* Joel Hanhart
Department of Ophthalmology, Shaare Zedek Medical Center, 12
Beyt Street, 91031 Jerusalem, Israel
Graefes Arch Clin Exp Ophthalmol (2017) 255:1859–1860